On Sat, 15 Jan 2005 03:16:11 -0500, Pete Smith wrote:
>I had fairly
>severe AC hum on the audio, apparently the result of a few millivolts
>potential difference between chassis. I finally cured it by experimentally
>connecting the chassis of the computer and radios directly to one another
>with heavy copper (not through the bus bar).
You were experiencing a "pin 1 problem," probably in both the computer and the
radio. Nearly all computer sound cards are built with pin 1 problems, and so
are nearly all ham rigs. The correct connection of a cable shield is to the
CHASSIS of the equipment on either end, NOT to the printed circuit board (often
called "audio ground" or "control ground" on the mic connectors and the rear
panel jacks. One sure sign of a pin 1 problem is an RCA or 1/8" jack that is
insulated from the chassis. The easy cure is to remove the shield from that
insulated jack and tie the shield to the chassis. Simply shorting the jack to
the chassis may establish ground loops interior to the equipment that are
problematic.
See http://audiosystemsgroup.com/Pin_1_Revisited.pdf
and http://audiosystemsgroup.com/Pin_1_Revisited_Part_2.pdf
While these applications notes specifically address balanced audio circuits,
the physics is the same whether the connection is balanced or unbalanced. The
only difference is that with an unbalanced connection, the current also causes
an IR (or IZ) drop in the interconnecting cable shield, which adds to the noise
produced by the pin 1 problem. But in most real systems, the noise/hum/buzz/rf
coupling produced by the pin 1 problem is greater than that caused by the IR/IZ
drop.
When you bonded the equipment together with beefy copper, you reduced the
voltage between them, which in turn reduced the current flowing on the cable
shield, which reduced the drop on the shield and the current flowing into the
improper shield connection to the printed circuit board.
BTW -- that current results from the potential difference between the two ends
of the shield, both of which are "grounded" to the power system's equipment
ground. If the connected equipment is plugged into different outlets, there can
be considerable noise (from hundreds of mV to a few volts) between the two ends
of the shield. Where does it come from? One major source is the line bypass
capacitors that are part of noise filters -- those caps couple noise to
"ground" and there is IR (and IZ) drop in the ground conductor. Another source
is leakage current from power transformers. In the pro audio world, two common
techniques are used to minimize the potential difference between "grounds." The
one that would apply to a ham shack is to power all interconnected equipment
from the same outlet, or, if you need more than one circuit, from outlets that
share the same equipment ground conductor (the green wire). Now, the only
voltage between ends of the cable is what can be induced by antenna action or
from magnetic coupling of a strong field. And you can minimize both by
minimizing the loop area between the signal cable and "ground." That means
running the signal cable tightly next to the big copper that bonds the
equipment together.
BTW -- one way to get that strong magnetic field is to tie neutral to ground at
some point in addition to the power entrance. Double-bonded neutrals are BAD
NEWS, and, as Tom corrected himself, are to be avoided at all costs.
Jim Brown K9YC
_______________________________________________
See: http://www.mscomputer.com for "Self Supporting Towers", "Wireless Weather
Stations", and lot's more. Call Toll Free, 1-800-333-9041 with any questions
and ask for Sherman, W2FLA.
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|